Christos D. Katsis

Toward Emotion Recognition in Car-Racing Drivers: A Biosignal Processing Approach

In this paper, we present a methodology and a wearable system for the evaluation of the emotional states of car-racing drivers. The proposed approach performs an assessment of the emotional states using facial electromyograms, electrocardiogram, respiration, and electrodermal activity. The system consists of the following: 1) the multisensorial wearable module; 2) the centralized computing module; and 3) the systems interface. The system has been preliminary validated by using data obtained from ten subjects in simulated racing conditions. The emotional classes identified are high stress, low stress, disappointment, and euphoria. Support vector machines (SVMs) and adaptive neuro-fuzzy inference system (ANFIS) have been used for the classification. The overall classification rates achieved by using tenfold cross validation are 79.3% and 76.7% for the SVM and the ANFIS, respectively.

An integrated telemedicine platform for the assessment of affective physiological states.

AUBADE is an integrated platform built for the affective assessment of individuals. The system performs evaluation of the emotional state by classifying vectors of features extracted from: facial Electromyogram, Respiration, Electrodermal Activity and Electrocardiogram. The AUBADE system consists of: (a) a multisensorial wearable, (b) a data acquisition and wireless communication module, (c) a feature extraction module, (d) a 3D facial animation module which is used for the projection of the obtained data through a generic 3D face model; whereas the end-user will be able to view the facial expression of the subject in real time, (e) an intelligent emotion recognition module, and (f) the AUBADE databases where the acquired signals along with the subjects animation videos are saved. The system is designed to be applied to human subjects operating under extreme stress conditions, in particular car racing drivers, and also to patients suffering from neurological and psychological disorders. AUBADEs classification accuracy into five predefined emotional classes (high stress, low stress, disappointment, euphoria and neutral face) is 86.0%. The pilot system applications and components are being tested and evaluated on Maseratis car. racing drivers.

A novel method for automated EMG decomposition and MUAP classification

OBJECTIVE This paper proposes a novel method for the extraction and classification of individual motor unit action potentials (MUAPs) from intramuscular electromyographic signals. METHODOLOGY The proposed method automatically detects the number of template MUAP clusters and classifies them into normal, neuropathic or myopathic. It consists of three steps: (i) preprocessing of electromyogram (EMG) recordings, (ii) MUAP detection and clustering and (iii) MUAP classification. RESULTS The approach has been validated using a dataset of EMG recordings and an annotated collection of MUAPs. The correct identification rate for MUAP clustering is 93, 95 and 92% for normal, myopathic and neuropathic, respectively. Ninety-one percent of the superimposed MUAPs were correctly identified. The obtained accuracy for MUAP classification is about 86%. CONCLUSION The proposed method, apart from efficient EMG decomposition addresses automatic MUAP classification to neuropathic, myopathic or normal classes directly from raw EMG signals.

A User Independent, Biosignal Based, Emotion Recognition Method

A physiological signal based emotion recognition method, for the assessment of three emotional classes: happiness, disgustand fear, is presented. Our approach consists of four steps: (i) biosignal acquisition, (ii) biosignal preprocessing and feature extraction, (iii) feature selection and (iv) classification. The input signals are facial electromyograms, the electrocardiogram, the respiration and the electrodermal skin response. We have constructed a dataset which consists of 9 healthy subjects. Moreover we present preliminary results which indicate on average, accuracy rates of 0.48,0.68 and 0.69 for recognition of happiness, disgust and fear emotions, respectively.

An integrated system based on physiological signals for the assessment of affective states in patients with anxiety disorders

Abstract Anxiety disorders are psychiatric disorders characterized by a constant and abnormal anxiety that interferes with daily-life activities. Their high prevalence in the general population and the severe limitations they cause have drawn attention to the development of new and efficient strategies for their treatment. In this work we describe the INTREPID system which provides an innovative and intelligent solution for the monitoring of patients with anxiety disorders during therapeutic sessions. It recognizes an individuals affective state based on 5 pre-defined classes (relaxed, neutral, startled, apprehensive and very apprehensive), from physiological data collected via non-invasive technologies (blood volume pulse, heart rate, galvanic skin response and respiration). The system is validated using data obtained through an emotion elicitation experiment based on the International Affective Picture System. Four different classification algorithms are implemented (Artificial Neural Networks, Support Vector Machines, Random Forests and a Neuro-Fuzzy System). The overall classification accuracy achieved is 84.3%.

A reasoning-based framework for car driver’s stress prediction

In this work, we present a novel methodology based on a dynamic Bayesian network for the estimation of car drivers stress produced due to specific driving events. the proposed methodology monitors driverpsilas stress using selected biosignals and provides a probabilistic framework in order to infer the driving events resulting in stress level increase. We conducted a series of experiments under real driving conditions. The extracted results indicate a strong correlation between the level of the stress as reported by the driver and the outcome of our model.

Automatic discovery of the number of MUAP clusters and superimposed MUAP decomposition in electromyograms

A novel data driven method for needle EMG decomposition is presented. The method is capable of automatically detecting the number of MUAPs. Superimposed MUAPs are detected and decomposed automatically into their constituents. No a priori knowledge of the number of MUAPs is required. The method is evaluated using a dataset consisting of 8 normal, 8 suffering from myopathy and 7 suffering from neuropathy subjects. The success rate on finding the correct number of clusters was 95%, 89% and 80%, respectively.

Can Ants Predict Bankruptcy? A Comparison Of Ant Colony Systems To Other State-Of-The-Art Computational Methods

In the current work, we consider the applicability of Ant Colony Systems (ACS) to the bankruptcy prediction problem. ACS are nature-based algorithms that mimic the functions of live organisms to find the best performing solution. In our work, ACS are used for the extraction of classification rules for bankruptcy prediction. An experimental study was conducted in order to evaluate the performance of the system and identify well performing parameters. Results were compared to the performance obtained by state-of-the-art methods for classification, namely the Artificial Neural Networks, the Support Vector Machines, the Partial Decision Trees and the Fuzzy Lattice Reasoning. Comparison indicates the high performance of the ACS which is further supported by their ability to extract classification rules, thus offering interpretation of the prediction results. The latter is of great importance in the field of corporate distress where no unified theory on distress prediction exists. Most studies with distress prediction have focused on increasing the accuracy of the model and have not always paid attention to the model interpretation.

INTREPID, a biosignal-based system for the monitoring of patients with anxiety disorders

In this work, we present the concept, the architecture and the evaluation of the INTREPID system. INTREPID is an advanced monitoring system which optimally classifies an individuals emotional state based on 5 pre-defined emotional classes, (relaxed, neutral, startled, apprehensive and very apprehensive), through association of the information arising from specific biological signals (Blood Volume Pulse, Heart Rate, Galvanic Skin Response and Respiration). The system is utilized for the monitoring of patients with anxiety disorder during therapeutic sessions. It is validated using data obtained through an emotion elicitation experiment based on the International Affective Picture System. An overall classification ratio of 85% is obtained.

PERFORM: First steps in the assessment of patient motion status and support to treatment changes

In the current work, a system for the monitoring, assessment and management of patients with chronic movement disorders such as Parkinsons disease (PD) is presented. The so called PERFORM system consists of the patient and the healthcare center subsystem. PERFORM monitors patients motion status in daily activities, using a set of light wearable sensors. Based on the analysis of the acquired signals, PERFORM assesses PD symptoms and their severity, integrates patients demographic, clinical and history data and proposes treatment plans based on advanced data mining algorithms. In this work we present two main modules of PERFORM system, the tremor assessment module and the data miner module.